The challenge to the juror Parmley we think was correctly overruled. It was not based upon the ground that he had any bias or prejudice against the prisoner, or that' he had formed any opinion or received any impression of the prisoner’s guilt; but it had for its sole foundation the juror’s belief that the crime charged had been committed by some one. The argument in support of the challenge is, that inasmuch as the first step in the case for the prosecution was to prove that the alleged offense had actually been committed by some one, and this point might possibly prove to be the one of principal contest iu the case, a juror *72who had already made up his mind concerning it, could not possibly eonie to the trial of the case with that freedom from prejudicial impressions, and that readiness to receive and weigh with impartiality the evidence given, which ought to exist when the liberty of the citizen is at stake; in other words, that he could not possibly be indifferent on this important branch of the case. There is some force in this argument, and if it were practicable in all cases to select a jury composed of persons- who were entirely free from any impressions whatever upon any question which might become one of. dispute in the" case, it might be exceedingly desirable to lay down some rule which would require it. But the rules upon this subject must be reasonable, and must keep in mind, the main purposes to be accomplished by jury trial. An intelligent jury is as important to a correct conclusion as one free from preconceived impressions; but it would- be impossible in a great many cases to summon twelve men of intelligence, who Were in the habit of availing themselves of the information on public affairs to be obtained from the usual channels, who would at the same time be free from any opinion regarding the commission of the offense. To exclude persons from the jury in a case of homicide because from the newspaper accounts they believed a felonious' killing had taken place, or in a case of treason, because they had sufficient knowledge of public events to' know that a rebellion had existed, would be to take a long step towards making stupidity the test of legal, capacity.' The ruling of the circuit judge was justified by the decision in Holt v. People, 13 Mich., 224, to which we adhere.
Nor do we think the objection that the loss of the Chicago letter was not sufficiently established to admit of. the introduction of parol evidence of its contents was well taken. The letter was shown to have been intrusted to *73the prosecuting attorney, and- he had looked for it in such places as he believed he had reason to expect to find it in. What search is sufficient in these cases is a mixed question of law and- fact, upon which the opinion of the person who has made it has more or less weight, and some intendment should be made in favor of the ruling of the circuit judge, who can generally judge better of the witness, and- of the good faith and thoroughness of his examination for the missing paper, than it is- possible for us to do.
The most important question in this connection is not whether the loss of the letter was- sufficiently accounted for, but whether the letter itself, if produced, would have been competent evidence. The writer of the letter,- it appears, was the principal witness against the prisoner, and had testified to a conversation had with him in Chicago, in which- the prisoner made statements indicating' his- participation in this offense'. On his cross-examination he was asked whether- he did not have a conversation with Edward O’Connor, Eobert McKinney and Michael Kilduff, - one morning during the examination of the prisoner before the justice, in which he told said O’Connor, McKinney and Kilduff that the prisoner was not the man with whom he had the conversation at Chicago; and he replied- in the negative.- He made, however, the following statement: McKinney, Kilduff, and one Hamilton were his bail on a criminal charge pending' against him at the- time Stewart was arrested on his complaint, on the charge now being tried. O’Connor, Kilduff and McKinney came to his house and told him if he gave evidence against Stewart they would throw up his bail. He did give such evidence and was surrendered by his bail as they had threatened. These three persons were then called by the defense and testified-that the- witness did say to them at the time inquired about that the prisoner was not the man with whom he *74had the conversation in Chicago. In reply to this testimony, the prosecution claimed the right to put in evidence the letter in question, which was written by the witness in Chicago to his brother in Bay City, after the time of the-alleged conversation with the prisoner in Chicago, and which spoke of the prisoner being there, and said, “If you want him, send word.” The prosecution also offered to show by the jailor that before.'the prisoner was arrested, and before there was any talk of arresting him, the witness had made to him the same statement in regard to the conversation at Chicago which he had sworn to in court. The circuit judge admitted this evidence, and the defense-excepted.
The question upon this branch of the case appears to-be this:. Whether after an attempt to impeach a witness by showing that he has made out of court statements inconsistent with those sworn to, his evidence can be supported by the testimony of witnesses who show that on other occasions his account of the transaction has corresponded with that given in court. This question appears to us to be one of no ordinary difficulty.
If it were an established fact that the witness had made the contradictory statements, we should say that the supporting evidence here offered was not admissible. If a witness has given different accounts of an affair on several different occasions, the fact that he has repeated one of these accounts oftener than the opposite one, can scarcely be said to entitle it to any additional credence. A man untruthful out of court is not likely to be truthful in court and where the contradictory statements are proved, a jury is generally justified in rejecting the testimony of the witness altogether. But in these cases' the evidence of contradictory statements is not received until the witness has denied making them, so that an issue is always made *75between the witness sought to be impeached and the witness impeaching him. The jury, therefore, before they can determine how much the contradictory statements ought to shake the credit of the witness, are required first to find from conflicting evidence whether he made them or not; and the question we now are to decide is, whether upon an issue of this character evidence like that received by the circuit judge was admissible.
The proper test for the admissibility of evidence ought to be, we think, whether it has a tendency to affect belief in the mind of a reasonably cautious person, who should receive -and weigh it with judicial fairness. Now there are many cases in' which, if evidence is given of statements made by a witness in conflict with those he has sworn to, his previous statements should not only be received in support of his credit, but would tend very strongly in that direction. If, for instance, the witness is himself the prosecutor, and has already made sworn complaint, there could be no doubt, we suppose, that the pendency of this complaint, its contents and the relation of the witness to it, might be put in evidence, and that they would raise a strong probability that the testimony to conflicting accounts as having been given about the same time, was either mistaken or corrupt. Suppose a person to be testifying in a case in which he had spent a considerable period of time and- a large sum. of money in pursuing an alleged criminal to conviction, and he is confronted with evidence of his own conflicting statements; the rule would be exceedingly unjust, as well as unphilosophical, which should preclude his showing, at least by his own evidence, such circumstances of his connection with the case as would make the impeaching evidence appear to be at war with all the probabilities. And other cases may readily be supposed in which, under the peculiar circumstances, the fact that the witness has always *76previously given a consistent account of the transaction in question, might well be accepted by the jury as almost conclusive that he had not varied from it in the single instance testified to for the purposes of impeachment.
It is impossible to lay down any arbitrary rule which could be properly applied to every case in which this question could arise; but we think there • are some cases in which the peculiar circumstances would render this species of evidence important and forcible. The tender age of the principal witness might sometimes be an important consideration, and the fact that the previous statement was put in writing, — as it was in this instance, — at a time when it would be reasonably free from suspicion, might very well be a controlling circumstance. We think the circuit judge ought to be allowed a reasonable discretion in such cases, and that though such evidence should not generally be received, yet that his discretion in receiving it ought not to be set aside except in a clear case of abuse; such as, we think, did not exist here.
The court was clearly in error, we think, in declining to allow the jury to be polled. It was objected that this question was not open here, because no exception was taken; but we think that what took place was sufficient to save the prisoner’s rights. He demanded the right to poll the jury, and,- when the court denied it, he strenuously insisted and protested against the verdict being recorded without such polling. Specific exceptions' are required in order to show beyond question that the party intends no waiver, but fairly notifies the opposite party and the court that the objection is to be insisted upon. The notification was ample in this case and the circuit judge must have fully understood it, as otherwise it is not to be supposed he would have incorporated what took place on this subject in the bill of exceptions. It is to be borne in mind that *77all this took place at the conclusion of the trial, and after the time when exceptions are commonly taken, and more liberality may, therefore, be properly extended to the party in respect to the form of his exception than if it related to earlier proceedings. Considering it sufficient, we proceed to a discussion of the judge’s ruling. The ground of the refusal was that the prisoner had consented to a sealed verdict; and one having been returned accordingly, he was precluded by this consent from the inquiries which would have been admissible otherwise, But why this should preclude him is not very evident. There is certainly “nothing in the mere fact that a juror’s name is appended to a written conclusion which negatives the possibility that the conclusion is one he does not assent to. Hpw is the prisoner to know that this is really the verdict of all; that all have personally signed it, and no one has been coerced, cajoled or deceived into assenting to that which his judgment does not concur in ? There is no mode of trying a question of this description, except by taking the evidence of the jurors themselves when they return into court; the case cannot stop to await the result of a suit in chancery to set aside the verdict for fraud or oppression. It is not claimed by the prosecution, as we understand it, that the jurors are so far concluded by their signatures that no questions whatever are to be put to them afterwards, but it is insisted that the prisoner has no right to have the inquiry pushed beyond the formal question which is always put to the whole jury together, whether this is the verdict of all. •
In other cases it is conceded that the right to have the jury polled cannot be questioned; and experience shows that notwithstanding the formal response, there are many cases in which individual jurors refuse to assent on being polled. This formal response is consequently not evidence of a very *78conclusive character, and unless a juror is absolutely bound by his signature to the written verdict, or unless the signature is such conclusive evidence of his deliberate assent that all other inquiries must be regarded as frivolous and useless, we do not discover any safe ground for denying to the prisoner the same right to have the jury polled in these cases which is conceded to exist in all others.
That the juror is not estopped by his signature from refusing his assent to the verdict when it is returned in open court is very clear. There is nothing in the nature of an estoppel in the case, and if on more mature reflection he could not concur in the verdict he has a right to refuse to be bound at any time before it is finally recorded; and we think the, prisoner has the same right to have the question of assent put to each individual juror in case a sealed verdict is returned that he has in any other. The sealed verdict is assented to only as a matter of convenience, and he has no reason to understand that he thereby waives any substantial right, or deprives himself of the privilege of determining, on the most conclusive evidence, that the verdict the jurors return is one which has the solemn and deliberate assent of all.
■ The only remaining question we have occasion to consider is, whether the circuit judge erred in refusing to grant a new trial on the ground that, after the verdict, it was discovered that one of the jurors who was summoned as a talesman, was not named on any of the assessment rolls for the county, either for the year in which the trial took place, or for any preceding year. We think he. did not.
The defense appear to suppose that no one is a competent juror unless he is assessed for taxation. But we do not find that the statute makes this a necessary qualification. It is true that the selection of persons to be returned as *79jurors is to be made from the assessment rolls; but the qualifications required, are only those of electors, — S. L 1861, p. 21, — and when once returned, the names remain in the box until others are returned in their places, whether they appear upon the successive assessment rolls or not.— Oomp. L, § 4856. It is not, therefore, a necessary qualification of a juror that he be a tax-payer. See Peninsular Railway Co. v. Howard, 20 Mich., 26; and as talesmen are to be summoned from the bystanders in court, or from the neighboring citizens, there is no occasion for inquiring whether their names appear upon the assessment rolls or not.
For the error in denying to the prisoner the right to have the jury polled, the judgment must be reversed and a new trial ordered.
Graves, J., concurred.